glm with poisson errors

Dear R users,

I have data on percentage leaf area damaged (in classes, e.g. 1%, 5%, 
10%) in plants. My two questions are:

(1) Could I use a glm with poisson errors on these data?
(2) Could I still use this glm with poisson errors after arcsine 
transformation of the data?

Thank you very much for your help!

Best regards
Christoph